Α New Agent-based Solution for Bridge Lifespan Extension

Abstract

Bridges are one of the most important and useful components of the transportation infrastructure, as they are crucial to many aspects of daily life, particularly in terms of economic development. However, bridges are vulnerable to several damages, malfunctions, and collapses that may considerably reduce their lifespan. The main causes of bridge fatigue are congested areas, excessive traffic, overloads, and additional vibration brought on by vehicles crossing them over. These elements reduce a bridge's longevity and raise maintenance costs. For a bridge to last as long as possible, it must not be crossed by an excessive number of cars and big trucks at once, and traffic should be kept under control. This study proposes an intelligent and autonomous multiagent-based system that incorporates a deep learning model to improve the monitoring of a bridge's crossing traffic. The idea was to propose a novel system for monitoring autonomously crossing vehicles in real-time to prevent the hastening of a bridge's infrastructure decline. The proposed system aimed to (i) increase the safety of drivers; (ii) lower maintenance costs; (iii) lessen the likelihood of bridge collapse, and (iv) lengthen bridge lifespans. Using a sample dataset, a straightforward deep learning model (CNN) was tested to improve bridge traffic monitoring. The performance of the proposed model was compared with the VGG-19 model. The results showed that the proposed model was effective in determining traffic congestion status with a 95% accuracy. As a result, the proposed system can be deployed on any bridge and can reduce crossing traffic overloads, extending its lifespan.